Two articles last week illustrated perfectly a lesser-known wiki aphorism: don't generate wiki content. Many a wiki concept has started off with a large existing data set. The idea is that if the data set is converted into a wiki, users will build upon the data set and add value, transforming commodity data into something more valuable.

EditMe's professional services team has seen more than a few such projects - some more successful than others. But in most cases, the parade of edits and enhancements that was hoped for falls short of expectations.

The articles that illustrated this so well are about city wikis. One highly successful example of a city wiki is the Davis, California Wiki. Clicking through the site shows a lot of care and thought has gone into it. Clearly written by and for Davis natives, the site is an example of the richness an active wiki can bring to a topic. A truly great city wiki provides answers to insider questions like where can I karaoke on Tuesdays, where's the best graffiti art and what is that smell? 

The article that turned me onto the Davis wiki was Nieman Journalism Lab's Welcome to Davis, Calif.: Six Lessons from the World's Best Local Wiki. The first lesson is "Wikis need content to breed content" - an obvious statement, but with subtle conditions. An empty or even very thin wiki, no matter how compelling the topic or engaged the audience, is unlikely to grow into something big and vibrant. Contributors need to feel that there's enough investment established to warrant their additional time and input. The content that's already there will prompt additional content and expansion.

But taken to the extreme, this is not true. Pump a wiki full of data and you're likely to end up with a ghost town. Or as a post on Newsless.org puts it, "a wiki written primarily by robots will appeal primarily to robots". That quote is taken from a post about the WikiCity Project, which aims to "provide a wiki for every city". It does that, but the vast majority of them are empty. Is a few pages in a wiki generated from a database really a wiki?

The site's page for my town, Maynard MA, makes this clear. I go there eager to see my city's wiki only to find content generated by a database of statistics, a Google map, a picture of a "Welcome to Massachusetts" sign that I don't recognize, and content ripped from Wikipedia (several versions old, even). To make matters worse, there isn't just one page like this for my town, but several pages ranging from "All about" to a business directory to an empty community calendar. It's a nice idea, but clearly, this isn't drawing editors. This data is not useful but is, more than anything else, disheartening to look at. By comparison, Wikipedia sports a decent single page entry for Maynard with a photo that looks like it could have been taken from my office doorstep.

At the end of the day, generated content in a wiki needs to give hope, not despair, to those who find it. If your data is sufficient enough to be useful on its own as a non-wiki web site, it might work with some luck and careful introduction. But if your data fits into spreadsheet columns, skip it. Build the most important pages by hand, add some real content, invite some contributors, and grow it organically.